## Are you sure?

This action might not be possible to undo. Are you sure you want to continue?

com/locate/energy

**A risk based heat exchanger analysis subject to fouling Part I: Performance evaluation
**

Syed M. Zubair *, Anwar K. Sheikh, Muhammad Younas, M.O. Budair

Department of Mechanical Engineering, King Fahd University of Petroleum & Minerals, Dhahran 31261, Saudi Arabia

Abstract Heat exchangers operating in the power and process industries are fouled to a greater or lesser extent depending on surface temperature, surface condition, material of construction, ﬂuid velocity, ﬂow geometry and ﬂuid composition. This fouling phenomenon is time-dependent and will result in a decrease in the thermal effectiveness of a heat exchanger. Once the thermal effectiveness decreases to a minimum acceptable level, cleaning of the equipment becomes necessary to restore the performance. In this paper, we present a simple probabilistic approach to characterize various fouling models that are commonly encountered in many industrial processes. These random fouling growth models are then used to investigate the impact on risk-based thermal effectiveness, overall heat-transfer coefﬁcient and the hot- and cold-ﬂuid outlet temperatures of a shell-and-tube heat exchanger. All the results are presented in a generalized form in order to demonstrate the generality of the risk-based procedure discussed in this paper. © 2000 Elsevier Science Ltd. All rights reserved.

1. Introduction Heat exchangers are extensively used in the power and process industries to transfer heat from one ﬂuid stream to another. The thermal–hydraulic performance of these heat exchangers decreases continuously with time due to fouling, which is deﬁned as the formation of deposits on heat transfer equipment (HTE). These deposits may be due to sedimentation, crystallization, organic or biological growths, corrosion products, or a combination of these effects [1–5]. In addition, where the heat ﬂux is relatively high, as in steam generators, fouling can lead to local hot spots and ultimately may result in mechanical failure of HTE, and hence an unscheduled shutdown of the plant. This may result in both economic and human loss, particularly in reﬁneries and thermal power plants. It is also important to mention that the designers and operators of HTE

* Corresponding author. Tel.: +966-3-860-2540; fax: 966-3-860-2949. E-mail address: smzubair@kfupm.edu.sa (S.M. Zubair).

0360-5442/00/$ - see front matter © 2000 Elsevier Science Ltd. All rights reserved. PII: S 0 3 6 0 - 5 4 4 2 ( 9 9 ) 0 0 0 8 0 - 8

i h.c T t tC U heat transfer area (m2) ﬂuid capacitance rate (W K 1) ˙ ˙ ﬂuid capacitance ratio (Cr=Cmin/Cmax) 1 probability density function (h ) heat transfer equipment enthalpy change (J) median time (h) mass ﬂow rate (kg h 1) number of shell passes ˙ number of transfer units (NTU=UA/Cmin) probability risk level heat transfer rate (W) fouling resistance (m2 K W 1) critical fouling resistance (m2 K W 1) temperature (K) time (h) time constant (h 1) overall heat-transfer coefﬁcient (W m 2 K 1 ) Greek symbols √a e () f scatter in time heat exchanger effectiveness cumulative normal distribution function rate of deposition or removal (m2 K J 1) Subscripts C c c. / Energy 25 (2000) 427–443 Nomenclature A ˙ C Cr f HTE H M m ˙ n NTU P p ˙ Q Rf Rf.o clean condition cold ﬂuid cold-ﬂuid inlet cold-ﬂuid outlet deposition fouled condition fouling hot ﬂuid hot-ﬂuid inlet hot-ﬂuid outlet .M.428 S. Zubair et al.o d F f h h.i c.

Zubair et al.2. 2. In this paper. powerlaw.10] and Sheikh et al.M. / Energy 25 (2000) 427–443 429 max min r maximum minimum removal Superscripts ∗ n asymptotic value exponent for the power-law model must be able to predict performance variations as the fouling proceeds. dt (1) It should be noted that the rate of fouling deposition. organic material growth. falling-rate and asymptotic fouling growth models and are discussed by Zubair et al. while the user of HTE must be able to formulate rational operating schedules for equipment-management purposes. [11]. fr. crystallization. To answer some of the questions of the designers and operators of HTE. This time-dependent behavior of the fouling process can be characterized by a number of fouling models proposed in the literature [1. we present a risk-based (or probabilistic) approach to the analysis of fouling models and then describe its impact on the thermal performance of heat exchangers. Zubair et al. [9. fd.5–8]. as well as the system conﬁguration. Dhahran. In particular. One of the major objectives of the project is to provide a risk-based mathematical structure to the heat exchanger fouling problem and then discuss rational maintenance and/or replacement strategies to reduce the overall maintenance and operating costs of heat exchangers subject to fouling. The designer needs this information to ensure that the user requirements with regard to preventive maintenance schedules can be met.10] and Sheikh et al. depends on the type of fouling mechanism (sedimentation.). while the rate of fouling removal. [11] have demonstrated that these models are probabilistic in nature.S. The different models result from the large number of uncertainties associated with the fouling growth process. These models are the linear. Fouling growth models The most widely accepted fouling model is based on the following general material balance equation ﬁrst proposed by Kern and Seaton [6]: dRf(t) fd fr. [9. etc. The rates of deposition and removal have been given many different forms (depending upon the type and mechanism of fouling) by various investigators. a project is being carried out at the King Fahd University of Petroleum and Minerals. depends on both the hardness or adhesive strength of the deposit and the shear stress due to the ﬂow velocity. They have shown that there is a considerable . particularly when considering actual heat exchanger applications. the distribution of fouling resistance of heat exchanger tubes with respect to time characterizes these random processes.

2. / Energy 25 (2000) 427–443 scatter in Rf(t) at any time t and. when Rf(t) is a linear function of time.c[1 a −1 (p)](t/M).50=P(Rf(t) Rf. 1 shows the representation of Eq.20.e. Sheikh et al. i. 1. where Rf.430 S. [14] while investigating the mechanism of CaCO3 scale growth in laboratory experiments. [12]. there is a corresponding scatter in the values of t. similarly. [1].c) Rf. initially proposed by Taborek et al. The strong bonds characterizing the structure of such deposits prevent signiﬁcant removal rates. p=0. Reduced random fouling resistance versus reduced time for a linear model with different values of risk level p and scatter parameter: (a) √a=0.c(t/tp. (2) Fig. was further supported by the experimental work of Hasson et al. for any ﬁxed value of Rf..1. these two parameters have a signiﬁcant impact on the thermal performance of heat exchangers.e. In this regard. Zubair et al. [13] and Zubair et al.. This hypothesis. M. and (b) √a=0. a random linear fouling growth model for the case when the critical acceptable value of fouling resistance Rf. Linear fouling model Linear fouling behavior is generally associated with the crystallization of a well-formed deposit consisting of a substantially pure salt that is largely uncontaminated by the presence of co-precipitated impurities. p.M.c is the critical acceptable value of fouling resistance. .30. [11] have also demonstrated that for a ﬁxed risk level p=P(Rf(t) Rf. As will be demonstrated in the following sections. to reach the critical level of fouling [i. (2) in terms of reduced time t/M and risk level p for a Fig. the time to reach this critical value will be distributed as an alpha distribution for a linear fouling growth model. a) Rf. It is important to mention that the probability distribution function has two main indicators.c).c)] and a scatter in the growth rate of the fouling process described by a scatter parameter √a. namely the median time. Andritsos et al.c is given can be expressed in terms of risk level p and scatter parameter √a as Rf(t.

[11] have also shown that replicate corrosion fouling data of Somerscales and Kassemi [16] can be represented by a power-law model.c M/[1 a −1 (p)]1/n.e. [15] while investigating the deposition of CaCO3 under relatively high operating temperatures.M. 2 illustrates the fouling resistance as a function of reduced time t/M and the risk level p Fig.50). (b) in a nontransformed x-axis. while Fig. 1(b) shows somewhat more scatter and that an even smaller time is required to reach the critical value at a given risk level compared with Fig. / Energy 25 (2000) 427–443 431 given scatter parameter √a. 1(a). scatter parameter √a=0. Zubair et al. As expected. (3) where the risk-based time to reach the critical fouling resistance for the power-law model is expressed as tp. p=0.30 and power of the exponent n=0. Fig. a) Rf.c[1 a −1 (p)]tn/Mn. A random version of the power-law fouling growth model on a linearly transformed time scale can be expressed similarly to the linear fouling model as Rf(t. the time to reach the critical level of fouling is much smaller than for the deterministic case (i.30.20. 2.S. It can be seen from these ﬁgures that for a low risk level. 2. for example when p=0.c(t/tp. (4) It is important to mention that M and √a represent the median time and the scatter parameter. 1(a) is for √a=0. Fig. Power-law fouling model The power-law fouling growth mechanism has been observed by Khan et al. Reduced random fouling resistance versus reduced time for a power-law model with different values of risk level p.e.. . p.c)n Rf. Fig.50: (a) in a transformed x-axis.01 (i. 1(b) is for √a=0. 99% reliability). respectively.2. Sheikh et al. in a transformed coordinate system where the power-law model is treated as a linear model..

which is representative of corrosion fouling data presented in Ref.50.c[1 a −1 (p)]ln(t)/ln(M). In this regard.. we note that the scatter in the growth rate in the transformed coordinate system is different from that in the non-transformed coordinate system. (5) Here. 2(b). p=0. Falling-rate fouling model Falling-rate fouling normally occurs in situations where the deposition rate is always greater than the removal rate. 2(a) with Fig.e. In addition. the exponent for the power-law model is taken as n=0. 2. where the falling-rate model is treated as a linear model. As expected. . This type of fouling mechanism has been observed by Muller-Steinhagen et al. / Energy 25 (2000) 427–443 for the scatter parameter √a=0. [17] while investigating the inﬂuence of operating conditions on particulate fouling.30.M. On comparing Fig.3. for a low risk level. a) Rf. the time to reach the critical level of fouling is much smaller than for the deterministic case (i. [16]. p.30: (a) in a transformed x-axis. (b) in a non-transformed x-axis. the data of Bansal and Muller-Steinhagen [18] for crystallization fouling in a plate exchanger also show a falling-rate trend.c)] Rf. Zubair et al. In plotting this ﬁgure.50). the risk-based time for the falling-rate model is expressed as ln(tp. 3 shows the plot of reduced fouling resistance Rf(t)/Rf. 3.c[ln(t)/ln(tp. (6) Fig. Reduced random fouling resistance versus reduced time for a falling-rate model with different values of risk level p and scatter parameter √a=0. It should be noted that random falling-rate fouling data on a linearly transformed time scale could be expressed in a manner similar to that of the powerlaw model as Rf(t. also similar to the power law model.c as a function of reduced time t/M Fig.c) ln(M)/[1 a −1 (p)]. this ﬁgure also shows that. the parameters M and √a represent the median time and the scatter parameter in a linearly transformed coordinate system.432 S.

This critical value may be considered as a representative number for most cooling water heat exchangers when it becomes necessary to clean the exchanger. so that the removal processes become progressively more effective with the deposit thickness. .30. Fig. In these heat exchangers. Also. (7) and rearranging.e..c M/[1 a −1 (8) (p)]. the reduced time to reach the critical level for p=0.S. f (7) The above equation may be explained as a linear version of the asymptotic fouling growth model. the conditions leading to the formation of a scale layer of a weak.50) compared with a low risk level case. we get Rf(t.4.50). 2. 90% reliability) is about 30% that of the deterministic case. where the time constant tC is expressed in terms of the critical acceptable value of fouling resistance Rf.10 (i.. we notice that the time to reach the critical level of fouling is large for a deterministic case (p=0. Zubair et al. the parameters M and √a represent the median time and the scatter parameter in a transformed coordinate system. expressed as tC tp. in plotting this ﬁgure. Again.M. It should be noted that.c/R∗)][1 f f a −1 (p)]t/M}].c/R∗)] f and tp. / Energy 25 (2000) 427–443 433 and risk level p for the scatter parameter √a=0. the time to reach the critical level of fouling is faster than in the deterministic case (i. For example. Such considerations lead to an asymptotic scale thickness.30. we have considered the critical fouling resistance as 95% that of the asymptotic value. or with the presence of suspended particles embedded in the crystalline structure. In this ﬁgure we note again that the scatter in the fouling growth rate in the transformed coordinate system is different from that in the nontransformed coordinate system. p=0. Asymptotic fouling model An asymptotic fouling growth model is often observed in cooling water heat exchangers.c and the time to reach this critical value tp.c/ln[1/(1 Rf. we notice in this ﬁgure that the scatter in the growth rate is much different in the transformed coordinate system when compared with the non-transformed coordinate system. for a low risk level. at which the deposition is balanced by the scale removal mechanism.c. (10) Similar to the above two random fouling growth models. where the exponential fouling growth model may be treated as a linear model. less coherent structure are associated with the simultaneous crystallization of salts of different crystal shapes. The random (or probabilistic) asymptotic fouling growth model on a transformed y-axis can be written as ln[1/(1 Rf/R∗)] t/tC. Hasson [19] has indicated that growth of such deposits is expected to create internal stresses in the scale layer. this ﬁgure also shows that. p. As expected. (9) Substituting the value of tC into Eq.e. a) R∗[1 exp{ ln[1/(1 Rf. 4 shows the plot of fouling resistance as a function of reduced time t/M and the risk level p for the scatter parameter √a=0.

mc (kg h 1) ˙ Hot-side mass-ﬂow rate.78 478.40 3 2 325. Tc. M (days) Total heat transfer area of the heat exchanger.66 2. Rf. n Number of tube passes per shell Inlet temperature (cold stream).70 3 . Rf.95R∗. It normally decreases with time due to increased fouling resistance.20 401. For an asymptotic fouling model. f 100 1070 27. power-law and asymptotic fouling models. For the present risk-based thermal analysis of a heat exchanger.55×10 452.i (K) Inlet temperature (hot stream). 3. UC (W m 2 K 1) Critical fouling resistanceb. The relevant properties of the exchanger are shown in Table 1. Th. Th.o(0) (K) Initial outlet temperature (hot stream).30: (a) in a transformed y-axis. Thermal performance of a heat exchanger The thermal performance of a heat exchanger is deﬁned in terms of temperature effectiveness. mh (kg h 1) ˙ Overall heat transfer coefﬁcient. A (m2) ˙ Maximum heat transfer duty. we consider a shell-and-tube heat exchanger in a crude oil preheat train. Reduced random fouling resistance versus reduced time for an exponential model with different values of risk level p and scatter parameter √a=0. Zubair et al. (b) in a non-transformed y-axis.and cold-ﬂuid streams.00 424. 4.600 145. It should be emphasized that at any time ‘t’ corresponding to a targeted risk level p and known Table 1 Relevant properties for the ﬂuid streams in the heat exchanger 1 2 3 4 5 6 7 8 9 10 11 12 13 a b Median timea.M.c=0.i (K) Cold-side mass-ﬂow rate.922 230. / Energy 25 (2000) 427–443 Fig.o(0) (K) The value of M is not needed for linear. Tc. which is a function of the temperature difference of the hot. Qmax (MW) Number of shell passes.434 S.c (m2 K W 1) Initial outlet temperature (cold stream).

o a2T 2 a3T 3 )].i . a) ˙ Qmax {[1−e1(t. √a) is the effectiveness of a heat exchanger with one shell pass and 2. a) U(t.and cold-ﬂuid temperatures are time. and the ﬂuid capacitance ratio (Cr=Cmin/Cmax) as [21] 1+exp[−NTU(t.i h. p. p. √a) and Rf(t. (15) U(t.and cold-ﬂuid temperatures in the Lee–Kessler equations [22]. a)Cr]/[1−e1(t. p. (13) The NTU at any time ‘t’ in the above equation is given by ˙ NTU(t.i−Th. ˙ ˙ NTU(t.o (17) . a) [(a1Th. 6n. a)]}n−Cr . p. p. a)−Tc. when ‘max’ belongs to the hot stream and ‘min’ to the cold stream.i a2T 2 a3T 3 ) (a1Th.M. we get Cr Th. a) Tc. we get after some simpliﬁcation [23] ˙ Qmax mh ˙ en(t. p. p. etc. p. a) m H(t. h. a) UC/[1 UcRf(t.S. p. r 1−exp[−NTU(t. a)(1+C2)1/2] r e1(t. (11) where e1(t. p. a)]}n−1 {[1−e1(t. respectively. p. notice that e1(0. a) 2/ 1 Cr (1 C2)1/2 .and risk-dependent. (14) where the time. a)]. p. p. ˙ (16) After substitution of the hot. and UC is the overall heat transfer coefﬁcient under clean conditions. a)Cr]/[1−e1(t. / Energy 25 (2000) 427–443 435 M and √a. p.i h. It is given in terms of the time-dependent number of transfer units. 6.o(t. hot. p. p. √a). p. √a) are the overall heat-transfer coefﬁcient and fouling resistance at any time ‘t’ with an associated risk level p and scatter parameter √a. a)(1+C2)1/2] r (12) ˙ It should be noted that Cmin is related to the ﬂuid stream which deﬁnes the maximum heat transfer rate that will occur in a counter-ﬂow heat exchanger with an inﬁnite area. 4. a)A/Cmin. p.o(t. etc. a). For example. p. p. tube passes can be expressed as [20] ˙ Q(t. p. It should also be noted that the outlet. a) en(t.and risk-dependent overall heat-transfer coefﬁcient can be expressed by considering fouling as an additional resistance to heat transfer as U(t. p. These temperatures can be easily obtained by calculating their respective enthalpies from the following expression ˙ Q(t. p. √a) is the effectiveness of the heat exchanger at time t=0 when there is no fouling. 4n.o h. the effectiveness of an exchanger with ‘n’ shell passes and 2n. Zubair et al. tube passes.

In these ﬁgures. Reduced temperature effectiveness versus reduced time for a linear random fouling model with different values of risk level p and scatter parameter √a=0.i. Th. while for the other models it is not required to assume the value of median time in our calculations.o(t. 4. Zubair et al. are plotted for linear. respectively. a). It is important to mention that for the falling-rate model we have considered M=100 days. p. falling-rate and asymptotic random fouling growth models in determining the heat exchanger’s effectiveness.M.and risk-dependent heat exchanger effectiveness based on a linear random fouling growth model of the exchanger. a2 and a3 are the constants for a ﬂuid stream. 5.c.436 S. This program is further extended to incorporate power-law. falling rate and asymptotic fouling-growth models. / Energy 25 (2000) 427–443 where Th. Also for the asymptotic model. as well as to study the impact of fouling on the overall heat-transfer coefﬁcient and hotand cold-ﬂuid outlet temperatures.30. √a)/en(0) versus reduced time t/M plots for different risk levels p and scatter parameter √a=0. Results and discussion The time. the critical fouling resistance Rf. the temperature effectiveness under clean conditions. p.30. (18) a1. power-law. Fig. A computer program was developed by Sheikh et al. which can be determined by using the procedure discussed in API Data Book [22]. and other relevant properties are given in Table 1.o Th. The overall heat-transfer coefﬁcient under clean condition UC. and Th.and risk-dependent temperature effectiveness of the shell-and-tube heat exchanger discussed in Section 3 are presented in Figs. we have considered the critical fouling resistance as 95% that of the asymptotic value. . [23] as a part of a general algorithm to study the time.o are the hot-ﬂuid inlet and outlet temperatures. reduced temperature effectiveness en(t. 5–8 in a reduced coordinate system.

It should be noted that the behavior of en(t.S. Reduced temperature effectiveness versus reduced time for a power-law random fouling model with different values of risk level p. Reduced temperature effectiveness versus reduced time for a falling-rate random fouling model with different values of risk level p and scatter parameter √a=0.01). Fig.M. p. / Energy 25 (2000) 427–443 437 Fig. however. 7.50. there is about a 17% decrease in effectiveness for a linear fouling model compared with clean conditions when t/M=1. scatter parameter √a=0.50) under the same operating condition. 9–12 for all .30 and power of the exponent n=0. for a low risk level (p=0. the reduced temperature effectiveness has a non-linear trend. the effectiveness of the heat exchanger degrades signiﬁcantly with time indicating that. However. Zubair et al. The inﬂuence of fouling on the overall heat-transfer coefﬁcient is shown in Figs. representing somewhat similar behavior to that of the respective underlying random fouling model. for other random fouling growth models. the decrease in effectiveness is about 10%. for the deterministic case (p=0. 6.30.00. As expected. √a)/en(0) is approximately linear for t/M 1.

Reduced temperature effectiveness versus reduced time for an exponential random fouling model with different values of risk level p and scatter parameter √a=0. in the case of linear growth of fouling (refer to Fig. four cases of random fouling growth models. As expected.10. / Energy 25 (2000) 427–443 Fig..00. 8. when time is equal to the median time.e.50) shows less degradation in the cleanliness factor compared with other risk levels. 9) when t/M=1. the deterministic case (p=0. Zubair et al.30. Reduced overall heat-transfer coefﬁcient versus reduced time for a linear random fouling model with different values of risk level p and scatter parameter √a=0. however. the overall heat-transfer coefﬁcient degrades with respect to reduced time.30. there is about 35% degradation in the value of U compared with the clean condition at p=0. i. 9. Similar to the temperature effectiveness. Fig. it is important to mention that for a ﬁxed value of cleanliness factor U(t)/UC the time to reach this critical value will be different for different values of risk level. Similar trends are also noted for the other three fouling growth models.M. Similar to the . For example.438 S.

30 are shown in Figs.30 and power of the exponent n=0. The variations in the reduced hot. these curves also show trends somewhat similar to that of the respective underlying fouling growth model. indicating that there will be a lower heat transfer rate due to fouling compared with the deterministic case.30. power-law. scatter parameter √a=0.e. Zubair et al. 10. . Fig. the hot-ﬂuid temperature is high while the cold-ﬂuid outlet temperature is low.M.and cold-ﬂuid outlet temperatures versus reduced time t/M for different risk levels p and for scatter parameter √a=0.50. the ﬁgures show that for a low risk level (i. reduced temperature effectiveness curves discussed above. Reduced overall heat-transfer coefﬁcient versus reduced time for a falling-rate random fouling model with different values of risk level p and scatter parameter √a=0. falling-rate and asymptotic fouling growth models. 13–16 for linear. 11. / Energy 25 (2000) 427–443 439 Fig. Reduced overall heat-transfer coefﬁcient versus reduced time for a power-law random fouling model with different values of risk level p.S. As one would expect.. high reliability) when compared with the deterministic case.

/ Energy 25 (2000) 427–443 Fig. Fig. 1–4) discussed earlier in Section 2. Zubair et al. 13. these curves also show that trends in temperature variations are very similar to that of the respective underlying fouling growth curve (refer to Figs. Reduced overall heat-transfer coefﬁcient versus reduced time for an exponential random fouling model with different values of risk level p and scatter parameter √a=0. 5. Concluding remarks A probabilistic approach is discussed to characterize various fouling growth models in terms of the risk level p and scatter in the growth rate of the process √a.30. Reduced outlet temperature versus reduced time for a linear random fouling model with different values of risk level p and scatter parameter √a=0. The models investigated are . Similar to the reduced temperature effectiveness and cleanliness factor curves.M. and (b) cold ﬂuid.440 S. 12.30: (a) hot ﬂuid.

The non-linear models are transformed into a linear coordinate system so that the probability distribution may be assumed as an alpha distribution in the transformed coordinate system. and (b) cold ﬂuid. / Energy 25 (2000) 427–443 441 Fig.30 and power of the exponent n=0. All the results are presented in graphical form to demonstrate the impact of risk level in assessing the performance of a heat exchanger.50: (a) hot ﬂuid. Fig. 14. it is clearly demonstrated .S. For example. Zubair et al. 15. linear. Reduced outlet temperature versus reduced time for a falling-rate random fouling model with different values of risk level p and scatter parameter √a: (a) hot ﬂuid.M. falling-rate and asymptotic. scatter parameter √a=0. These random fouling growth models are then considered in the performance evaluation of a shell-and-tube heat exchanger in a crude oil preheat train to demonstrate the inﬂuence of risk level and scatter parameter on important thermal parameters of the heat exchangers. power-law. and (b) cold ﬂuid. Reduced outlet temperature versus reduced time for a power-law random fouling model with different values of risk level p.

442 S. Palen JW. Fouling: the major unresolved problem in heat transfer. here. an acceptable level of the overall heat-transfer coefﬁcient will primarily govern the cleaning strategy. Ritter RB. This cost-based maintenance strategy is presented in the companion paper.68(2):59–67. References [1] Taborek J. . Chem Eng Progr 1972. Acknowledgements The authors acknowledge the support provided by King Fahd University of Petroleum & Minerals through research project ME/Fouling/176. However. the decisions about maintenance can be optimized by developing cost as a function of reliability (or risk level) and then searching for a minimum costbased solution. heat exchangers are not in a network or in a critical system. plate-and-frame and other compact heat exchangers. Reduced outlet temperature versus reduced time for an exponential random fouling model with different values of risk level p and scatter parameter √a=0. in some situations. / Energy 25 (2000) 427–443 Fig. which can often result in increased operation and maintenance costs. that increasing the risk level for a given critical fouling thermal resistance increases the performance characteristics of the exchanger. Although the analysis presented in the paper is applicable to shell-and-tube heat exchangers. the procedure can easily be modiﬁed to include other types of heat exchanger such as double pipe. for situations in which the cost of operation and maintenance is an important factor along with the exchanger reliability. Knudsen JG. In this regard. It is important to emphasize that operating a heat exchanger at a critical risk level is important in some applications such as a heat exchanger network in a reﬁnery. Aoki T.M. Zubair et al. and (b) cold ﬂuid.30: (a) hot ﬂuid. It is thus important that. maintaining the exchanger at a higher reliability level (or at a lower risk level) implies more frequent maintenance intervals. 16.

74:911–9. 1995. [22] Anon. Panchal CB. Haq MU. Watkinson AP. / Energy 25 (2000) 427–443 443 [2] Taborek J. Reliability-based maintenance strategies for heat exchangers subject to fouling. Washington (DC): Hemisphere. Ritter RB. Aoki T. Riyadh. Exchanger stages solved by graph. Budair MO. New York: Begell House. Knudsen JG. [17] Muller-Steinhagen H. Zubair SM. A maintenance strategy for heat-transfer equipment subject to fouling: a probabilistic approach. American Petroleum Institute.118(4):306–12. Procedure 7B4. [6] Kern DQ.115(3):584–91. Panchal CB. [15] Khan MS. Amsterdam: Elsevier Science Publishers Ltd. [12] Hasson D. 1999:393–400. Ashiru OA. Knudsen JG. In: Bott TR. Understanding Heat Exchanger Fouling and Its Mitigation. 1978. [10] Zubair SM. Fouling of heat transfer equipment. Somerscales EFC. Can J Chem Eng 1996. In: Proceedings of the 3rd Saudi Engineering Conference.17(8):769–76.66(1):42–50. Fouling of heat exchangers.49(7):137–45. Calcium carbonate deposit formation under isothermal conditions. Mechanisms of calcium carbonate deposition on heat transfer surfaces. Br Chem Eng 1959. Vol. .S. Muller-Steinhagen H. New York: Begell House. Heat exchanger design for fouling service. [7] Epstein N. Reif F. Understanding Heat Exchanger Fouling and Its Mitigation. Budair MO. Statistical aspects of CaCO3 fouling in AISI 316 stainless steel tubes. editors. King Saud University Press College of Engineering. Warner WJ. [20] Jamin B.4(5):258–62. Budair MO. 1991:1–11. Windreich S. Rozenman T. [16] Somerscales EFC. 4th edition. Avriel M. Budair MO.62(7):51–6.7(1):59–65. Inﬂuence of operating conditions on particulate fouling. Chem Eng Progr 1966. 1.119(3):581–8. Heat Mass Transf 1996. Fouling due to corrosion products formed on a heat-transfer surface. Ind Eng Chem Fundamentals 1968. In: Bott TR. Sheikh AK. Haq MU. Saudi Arabia. Younas M.32(1-2):73–80. Progress in precipitation fouling research — a review. Kassemi M.134. Somerscales EFC. Some remarks on the probabilistic basis for characterizing fouling data. [21] Kern DQ. Chem Eng Progr 1972. editors. Fouling: technical aspects (afterward to fouling in heat exchangers).55:374–80. [5] Bott TR. Resnick W. 1999:67–89. Can J Chem Eng 1976.M. Abdelaleem FA. Karabelas AJ.7: alternate computer method for enthalpy of petroleum fractions. Ritter RB. ASME J Heat Transf 1987. In: Somerscales EFC. Budair MO. Zubair SM. [4] Chenoweth JM. API data book — petroleum reﬁning. [3] Suitor JW. Shaik MN. [11] Sheikh AK. editors. Sheikh AK. [8] Al-Ahmad MI. Melo LF.109(1):267–71. Crystallization fouling in plate heat exchangers. A probabilistic approach to the maintenance of heat-transfer equipment subject to fouling. [19] Hasson D. Melo LF.11(1):73–107. [18] Bansal B. Zubair SM. 1981:31–53. Can J Chem Eng 1988. Sheikh AK. ASME J Heat Transf 1997. Zubair et al. Seaton RE. Epstein N. ASME J Heat Transf 1997. [13] Andritsos N. 7. Predictive methods for fouling behavior. Kontopoulou M. Quddus A. The history and status of research in fouling of exchangers in cooling water service. New York [23] Sheikh AK. Quddus A.119(3):575–80. Sheikh AK. ASME J Heat Transf 1993. Heat Transf Eng 1990. ASME J Energy Resourc Technol 1996. [14] Zubair SM. Fouling resistance model for prediction of CaCO3 scaling in AISI 316 tubes. Scale and fouling problems and its reﬂection on the performance of industrial thermal processes. Hydrocarbon Process 1970. Energy 1992. Badar MA.133-7. Palen JW. [9] Zubair SM. Final report of the HTRI/TEMA joint committee to review the fouling section of the TEMA standards. A theoretical analysis of thermal surface fouling.68(7):69–78.

- Development of a Condenser for Marine Florae Pyrolysis Reactor
- A Model for the Analysis of Compact Heat Exchangers
- SHMT sumair 1
- Heat Transfer ower
- Heat Exchnger Ttechnology-twisted
- A Conjetura on the Distribution of Firm Profit
- 2196.pdf
- Specifying Shell and Tube Heat Exchangers
- SP-210-E-103 REV 0
- phthal1
- Ccc
- Heat Exchanger
- Heat Transfer Performance Analysis of Water Nano Particles Combination for Corrugated Plate Heat Exchanger
- day-2.pptx
- Performance Analyses of Counter-Flow Closed Wet Cooling Towers Based on a Simplified Calculation Method
- Press Release HX Management Change Eng
- Heat Transfer Enhancement by using Twisted Tape Insert
- Flyer Diesta Web
- Full Report - Draft V5
- Integrated Distillation System With Prefractio
- Shell and Tube HEX Problem
- KOBELCO-Copper-Alloy Tubes for Heat Exchangers
- 1 st batch.pptx
- Heat Exchanger Design
- Design and Optimization of a Condenser for HVAC Automobile System.
- Anti-Vibration Tube Support Technologies - Abstract by Cpv
- Condenser
- test 13
- CH3080 Boilers and Condensers
- Automatic Tube Brushing system

- List of Commonly Used American Code and Standard in Process Industries
- TEMA Data Sheet
- Cone Development
- m398content.pdf
- Conv
- Compliance for SECTION 15540 Hot Water Boiler1
- Boiler Conversion
- ARABIC TO INDIA.xlsx
- The.nice.Guys.2016.1080p.web DL.srt
- 8.5 Welding Procedure
- HpRegulatorKullanımKlavuzu.pdf
- HpRegulatorKullanımKlavuzu.pdf
- HpRegulatorKullanımKlavuzu.pdf
- wc-blank
- ARABIC TO INDIA.xlsx
- Material Specifications
- 84052566-DMP-HTRI-SCH
- Leave Requisition Form 2 (3)
- Zootopia.2016.BDRip.x264 SPARKS.srt
- SAEP-334
- HpRegulatorKullanımKlavuzu
- Heat Exchangers
- Chemline Putty
- Material Specifications
- Impeller
- New in 2009
- New in 2009
- calc.txt
- வெண்டைக்காய் தயிர் பச்சடி - techsatish
- Print Itinerary_Gulf Air

Sign up to vote on this title

UsefulNot usefulClose Dialog## Are you sure?

This action might not be possible to undo. Are you sure you want to continue?

Loading